"by the larger eye and shorter thickset figure." 

 Pectoral fins described as reaching end of second 

 dorsal fin, longer than in T. albacares (as Thijnnus 

 albacora). 



Thy7inus sibiTcmm'mck and SMcgcl, 1844. Lec- 

 totype, RMNH 2327 (a mounted skin, 600 mm. fork 

 length), and paralectotype, RMNH 799 (right side 

 of mounted skin, backed bj' cardboard, 5o7 mm. fork 

 length) designated by Boeseman (1947). Measure- 

 ments made l)y Gibbs in 1902 fall in tlie range of T. 

 albacares rather than T. ohcsus, but mounted speci- 

 mens could be expected to be unreliable for this pur- 

 po.se. The description by Tcmminck and Schlegel 

 likewise offers little aid in identifying the species. 

 They note that the pectoral fin is shorter than in T. 

 alalunga (as Thynnus alnlonga or T. pacijicns) and 

 approaches in length that of T. albacares (as T. 

 argcnlivittalus), and their illustration shows a pec- 

 toral fin resembling that of a fairly small Pacific T. 

 obcxus. On this basis, we follow other authors in 

 considering T. sibi a synonym of T. obcsus. If the 

 measurements of the lectotype and paralectotype 

 were taken at face value, T. sibi would have to be 

 regarded as a synonym of T. albacares, but we prefer 

 for the present to disregard these specimens. 



Thun/ms mcbachi Kishinouyc, 1915. No t.ypo 

 specimens. Original description clearly referable to 

 T. obesus, apparently based on a number of 

 specimens. 



Characters 



Pectoral fin intermediate in length (22-31 percent 

 of fork length) in specimens longer than 1,100 mm. 

 (as in T. albacares and T. allanticxs), as long as in 

 T. alalunga (greater than 31 percent) in specimens 

 less than 1,100 mm. from the Pacific. 



Gill rakers 23-31 (generally similar to T. albacares 

 and T. alalunga). 



Liver with striations on ventral surface (not 

 restricted to margins, fig. 30), its three lobes subeciual 

 in length, vascular cones present on its dorsal side 

 (as in T. thynnus, T. maccoyii, and T. alalunga). 

 Spleen on right side, stomach on left (as in all except 

 T. alalunga). Swimbladder as long as body cavity, 

 with two globular dorsal heads anteriorly, tapering 

 gradually to a jwint posteriorly. Kidney with a 

 short tail, reaching the level of vertebra 11-13. 



Cutaneous artery usually originating at level of 

 vertebra 0-8, passing laterally between ribs 5 and G, 

 branching between intermuscular bones G anil 7 (as 

 in T. albacares, T. tonggol, and T. atlanlicus). Two 



rows of arterioles and venides arising from each main 

 lateral cutaneous branch (as in T. thynnus aiid T. 

 maccoyii). Post-cardinal vein present, joining right 

 cutaneous vein (as in 7'. albacares, T. tonggol, and 

 T. atlanlicus). 



Posterior parasphcnoid margin either rounded (in 

 some small specimens) or forming a slightly olituse 

 angle (not as acute as in T. alalunga, T. maccoyii, or 

 T. thynnus). 



Vertebrae 18 + 21 (as in all Thunnus except T. 

 atlanlicus). First \-ent rally directed parapo|)hysis 

 on vertebra 9 (as in all excciit T. tonggol ami T. 

 tliynnu.<i). First closed haemal arch usually on ver- 

 tebra 1 1 (as in T. albacares, T. allnnlicus, T. tonggol, 

 and some T. thynnus). Haemal i)rezygapophyses 

 ai'ising high on haemal arch (as in T. thynnus and T. 

 maccoyii). All haemal postzygapopiiyses short, less 

 than half centrum length (as in 7'. alalunga, T. 

 thynnus, and T. maccoyii). Antt-riorinost ventro- 

 lateral foramina small, not more tiian I'j times 

 width of haemal spine (as in T. alalunga, T. thynnus, 

 and T. maccoyii). 



Nominal species 



Three names have been applied to this species: 

 Thunnus obesus for the Atlantic poi)ulation and T. 

 ■nbi and T. meJ)achi for the Pacific population, both 

 latter names based on Japanese specimens. Fraser- 

 Hrunner (1950) correctly placeil sibi and mcbachi in 

 the synonymy of T. obesus but also mistakenly in- 

 cluded Thunnus maccoyii, T. phillipsi {=T. mac- 

 coyii), and Parath)innus rosengarteni (= T. atlanlicus) 

 as was pointed out by Rivas (19G1). Jones and 

 Silas (19G0) stated that they could find no notable 

 differences between Atlantic and Pacific populations, 

 but referred to their Indian Ocean specimens as 

 Paralhunnus obesus mcbachi. Because the name 

 sibi has priority over mcbachi, this should be T. 

 obesus sibi if tlie Indo-Pacitic population is subspe- 

 cifically differentiated, and the latter name was u.sed 

 bj- Jones and Silas (19G3). 



Rivas (I9G1) claimed that the Pacific populations 

 {"sibi") can be distinguished from the Atlantic T. 

 obesus at lengths of about a meter by a much longer 

 pectoral fin, but admitted difficulty in differentiating 

 large specimens. His conclusion was based on a 

 single small Atlantic specimen (74G mm., pectoral 

 29.4 percent of fork length) and 10 small specimens 

 (data from Dung and Pvoyce, 1953: 74) from the 

 western Marshall Islands (G0()-835 mm., pectoral 

 38.8-44.9 percent). 



110 



U.S. FISH AND WILIJLIKE SKltVICK 



